基于专利数据挖掘的技术融合识别与技术机会预测研究——以电动汽车产业为例

图书馆杂志 ›› 2019, Vol. 38 ›› Issue (10): 95-106.

基于专利数据挖掘的技术融合识别与技术机会预测研究——以电动汽车产业为例

慎金花闫倩倩孙乔宣万召侗#br#

出版日期:2019-10-15 发布日期:2019-10-29
作者简介:慎金花女，同济大学图书馆，党委书记，研究馆员。研究方向：情报理论与方法、信息管理与信息系统。作者贡献：参与研究方向、理论基础及研究方法的确立，参与论文撰写。E-mail：jhshen@lib. tongji.edu.cn 上海 200082 闫倩倩女，同济大学经济与管理学院，硕士研究生。研究方向：情报理论与方法、信息管理与信息系统。作者贡献：参与研究方向、理论基础及研究方法的确立，参与论文撰写。上海 200082 孙乔宣女，同济大学经济与管理学院，硕士研究生。研究方向：情报理论与方法、信息管理与信息系统。作者贡献：参与论文撰写与修订。上海 200082 万召侗同济大学图书馆，副研究馆员。研究方向：图书馆信息检索服务、计量分析。作者贡献：参与论文撰写与修改。上海 200082

Research on Technology Fusion Recognition and Technology Opportunity Prediction Based on Patent Data Mining

Shen Jinhua,Yan Qianqian, Sun Qiaoxuan, Wan Zhaodong

Online:2019-10-15 Published:2019-10-29

摘要/Abstract

摘要： 从技术融合的角度识别产业未来的技术机会可以帮助企业及时调整研发战略，更加高效
地开展研发工作。因此，本文基于专利数据挖掘技术构建了一种技术融合识别预测模型。模型通
过Apriori算法构造IPC共现网络，应用Louvain聚类算法将网络划分为不同的技术社群，并基于LDA
主题模型发掘各社群的技术主题，通过构建“社群融合潜能”和“技术融合价值”两项指标评估
了社群弱关系的融合价值，并将融合价值较高的弱关系组成技术融合预测网络，通过提取主题词
来发掘未来的技术机会。本文以电动汽车产业为例对该模型进行了实证。

关键词: 专利数据挖掘, 技术融合, IPC共现网络, IPC社群, 弱关系

Abstract: Identifying the future technology opportunities from the perspective of technology convergence
can help enterprises adjust their R&D strategies in time and carry out R&D activities more efficiently.
Therefore, based on the patent data mining technology, this paper constructs a technology fusion recognition
and prediction model. This model constructs IPC co-occurrence network through Apriori algorithm, splits
the network into different technology cluster by Louvain algorithm, and explores the technical themes of
each cluster using LDA topic model. By constructing two indicators, “community fusion potential” and
“technology fusion value”, we evaluate the fusion value of the weak ties between communities. The weak
ties with high fusion value will be assembled to form a prediction network, and the future technological
opportunities can be explored by extracting keywords from the prediction network. Taking the electric
vehicle industry as an example, this paper makes an empirical study on the model.

Key words: Patent data mining, Technology fusion, IPC co-occurrence network, IPC cluster, Weak ties

慎金花闫倩倩孙乔宣万召侗. 基于专利数据挖掘的技术融合识别与技术机会预测研究——以电动汽车产业为例[J]. 图书馆杂志, 2019, 38(10): 95-106.

.Shen Jinhua, Yan Qianqian, Sun Qiaoxuan, . Research on Technology Fusion Recognition and Technology Opportunity Prediction Based on Patent Data Mining[J]. Libraly Journal, 2019, 38(10): 95-106.

参考文献

[ 1 ] Lei D T. Industry Evolution and CompetenceDevelopment: the Imperatives of TechnologicalConergence[J]. International Journal of TechnologyManagement, 2000, 19(7-8): 699-738.
[ 2 ] Kodama F. Emerging Patterns of Innovation: Sourcesof Japan’s Technological Edge[J]. R&D Management,1996, 26(2): 179-181.
[ 3 ] Bores C, Saurina C, Torres R. TechnologicalConvergence: A Strategic Perspective[J].Technovation, 2003, 23(1): 1-13.
[ 4 ] 周磊, 杨威, 王芮. 技术融合视角下新兴技术识别研究[J]. 科技进步与对策, 2018, 35(15): 57-62.
[ 5 ] 周磊, 杨威. 竞争情报视角下突破性创新的识别思路[J]. 情报杂志, 2015, 34(1): 32-37.
[ 6 ] Kim B, Gazzola G, Yang J, et al. Two-phase edgeoutlier detection method for technology opportunitydiscovery[J]. Scientometrics, 2017, 113(1):1-16.
[ 7 ] Kim B, Gazzola G, Lee J M, et al. Inter-clusterconnectivity analysis for technology opportunitydiscovery[J]. Scientometrics, 2014, 98(3):1811-1825.
[ 8 ] Kim J, Lee S. Forecasting and identifying multitechnologyconvergence based on patent data:the case of IT and BT industries in 2020[J].Scientometrics, 2017, 111(1): 47-65.
[ 9 ] 江屏, 王川, 孙建广, 等. IPC聚类分析与TRIZ相结合的专利群规避设计方法与应用[J]. 机械工程学报, 2015(7): 144-154.
[10] 周磊, 杨威, 张玉峰. 基于专利挖掘的突破性创新识别框架研究[J]. 情报理论与实践, 2016, 39(9):73-76.
[11] Caviggioli F. Technology fusion: Identification andanalysis of the drivers of technology convergenceusing patent data[J]. Technovation, 2016, 55(56):22-32.
[12] Peter Harrington. 机器学习实战[M]. 北京: 人民邮电出版社, 2013: 201-209.
[13] 黄鲁成, 黄斌, 吴菲菲, 等. 基于专利共类的信息与生物技术融合趋势分析[J]. 情报杂志, 2014,33(8): 59-63.
[14] 马瑞琼. 复杂网络中社团发现算法的研究[D]. 成都: 电子科技大学, 2015.
[15] 勒孚刚. 基于LDA模型的专利文本分类及演化研究[D]. 赣州: 江西理工大学, 2017.
[16] Newman M E J. Fast algorithm for detectingcommunity structure in networks. [J]. Phys RevE Stat Nonlin Soft Matter Phys, 2003, 69(6 Pt 2):066133.
[17] Newman M E J, Girvan M. Finding and evaluatingcommunity structure in networks[J]. Physical ReviewE Statistical Nonlinear & Soft Matter Physics, 2004,69(2): 026113.
[18] Blondel V D, Guillaume J L, Lambiotte R, et al.Fast unfolding of communities in large networks[J].Journal of Statistical Mechanics, 2008(10): 155-168.
[19] Blei D M, Ng A Y, Jordan M I. Latent dirichletallocation[J].Machine Learning Research Archive, 2003(3): 993-1022.
[20] Granovetter M S. The Strength of Weak Ties[J].American Journal of Sociology, 1973, 78(6):1360-1380.
[21] 刘军. 整体网分析: UCINET软件实用指南[M]. 上海: 上海人民出版社, 2009: 126-136.
[22] De Meo P, Ferrara E, Fiumara G, et al. On Facebook,most ties are weak[J]. Communications of the Acm,2012, 57(11): 78-84.
[23] Asur S. An event-based framework for characterizingthe evolution of interaction graphs[J]. AcmTransactions on Knowledge Discovery from Data,2009, 3(4): 1-36.
[24] Knoke D, Song Y. 社会网络分析[M]. 上海: 上海人民出版社, 2017: 116-121.
[25] Freeman L C. A set of measures of centrality basedon betweenness. [J]. Sociometry, 1977, 40(1): 35-41.
[26] 百度百科. 百度阿波罗平台[OL]. https://baike.baidu.com/item/阿波罗/20625862.